Accepted Manuscript

Minimized natural versions of fungal ribotoxins show improved active site plasticity

Moisés Maestro-López, Miriam Olombrada, Lucía García-Ortega, Daniel Serrano-González, Javier Lacadena, Mercedes Oñaderra, José G. Gavilanes, ÁlvaroMartínez-del-Pozo

PII: S0003-9861(17)30034-6

DOI: 10.1016/j.abb.2017.03.002

Reference: YABBI 7446

To appear in: Archives of Biochemistry and Biophysics

Received Date: 18 January 2017

Revised Date: 3 March 2017

Accepted Date: 5 March 2017

Please cite this article as: M. Maestro-López, M. Olombrada, L. García-Ortega, D. Serrano-González,J. Lacadena, M. Oñaderra, J.G. Gavilanes, E. Martínez-del-Pozo, Minimized natural versions of fungalribotoxins show improved active site plasticity, Archives of Biochemistry and Biophysics (2017), doi:10.1016/j.abb.2017.03.002.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service toour customers we are providing this early version of the manuscript. The manuscript will undergocopyediting, typesetting, and review of the resulting proof before it is published in its final form. Pleasenote that during the production process errors may be discovered which could affect the content, and alllegal disclaimers that apply to the journal pertain.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

1

Minimized natural versions of fungal ribotoxins show improved active site plasticity 1

Moisés Maestro-López‡, Miriam Olombrada†, Lucía García-Ortega, Daniel Serrano-2

González, Javier Lacadena, Mercedes Oñaderra, José G. Gavilanes and Álvaro Martínez-3

del-Pozo* 4

Departamento de Bioquímica y Biología Molecular I, Facultad de Ciencias Químicas, 5

Universidad Complutense, 28040 Madrid, Spain. 6

†Present address: Department of Evolutionary Biology and Environmental Studies, 7

University of Zurich, Winterthurerstrasse 190, 8057 Zurich, Switzerland. 8

‡Present address: Departamento de Estructura de Macromoléculas, Centro Nacional de 9

Biotecnología (CNB-CSIC), Darwin 3, 28049 Madrid, Spain 10

*Corresponding author: Tel: 34 91 394 4259; E.mail: alvaromp@quim.ucm.es 11

Abstract 12

Fungal ribotoxins are highly specific extracellular RNases which cleave a single 13

phosphodiester bond at the ribosomal sarcin-ricin loop, inhibiting protein biosynthesis by 14

interfering with elongation factors. Most ribotoxins show high degree of conservation, with 15

similar sizes and amino acid sequence identities above 85%. Only two exceptions are 16

known: Hirsutellin A and anisoplin, produced by the entomopathogenic fungi Hirsutella 17

thompsonii and Metarhizium anisopliae, respectively. Both proteins are similar but smaller 18

than the other known ribotoxins (130 vs 150 amino acids), displaying only about 25% 19

sequence identity with them. They can be considered minimized natural versions of their 20

larger counterparts, best represented by α-sarcin. The conserved α-sarcin active site residue 21

Tyr48 has been replaced by the geometrically equivalent Asp, present in the minimized 22

ribotoxins, to produce and characterize the corresponding mutant. As a control, the inverse 23

anisoplin mutant (D43Y) has been also studied. The results show how the smaller versions 24

of ribotoxins represent an optimum compromise among conformational freedom, stability, 25

specificity, and active-site plasticity which allow these toxic proteins to accommodate the 26

characteristic abilities of ribotoxins into a shorter amino acid sequence and more stable 27

structure of intermediate size between that of other nontoxic fungal RNases and previously 28

known larger ribotoxins 29

Keywords: RNases; insecticidal; sarcin; hirsutellin; anisoplin 30

Abbreviations: CD, circular dichroism; HtA, Hirsutellin A; PDB, Protein Data Bank; 31

SEM, standard error of the mean; SRL, sarcin-ricin loop; WT, wild-type, Tm, melting 32

temperature. 33

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

2

Fungal ribotoxins are a unique group of highly specific extracellular RNases [1, 2] initially 34

discovered as antitumoral agents [3]. The toxicity of these ribotoxins relies on their ability 35

to cleave a singular phosphodiester bond strategically located at a universally conserved 36

sequence of the large rRNA [4], known as the sarcin-ricin loop (SRL). Therefore, 37

ribosomes are their natural substrates. Cleavage of this single bond inhibits protein 38

biosynthesis, interfering with the function of elongation factors [5-7] and leading to cell 39

death by apoptosis [8]. Given the universal conservation of the SRL sequence, all known 40

ribosomes are susceptible to ribotoxins action. However, these toxic RNases show different 41

affinity against ribosomes from different origins and, above all, they first have to cross the 42

plasmatic membrane to be able to cleave their substrate and exert their cytotoxic activity [9-43

11]. This behavior explains why not all cells are equally targeted by these toxins. 44

Intriguingly, they are especially active on transformed or virus-infected cells [3, 8, 12] due 45

to an altered permeability of their membrane in combination with an enrichment in acidic 46

phospholipids [8, 11, 13-15]. This feature has been lately related to the possibility of 47

ribotoxins acting as natural insecticidal agents [16-18]. 48

α-Sarcin, restrictocin, Aspf1, and hirsutellin A (HtA) are the most exhaustively 49

characterized ribotoxins [10, 11, 19-26], but many others have been identified and partially 50

characterized in different fungal species [27-35]. Most of them show a high degree of 51

conservation, with similar sizes and amino acid sequence identities above 85% [1, 18, 33]. 52

So far only two exceptions are known: HtA [23, 24, 32] and the recently discovered 53

anisoplin [35], produced by the entomopathogenic fungi Hirsutella thompsonii and 54

Metarhizium anisopliae, respectively. Both proteins are highly similar but smaller than the 55

other known ribotoxins (130 residues vs 150), displaying only about 25% sequence identity 56

with them [23, 24, 32, 35]. Therefore, they can be considered as minimized natural versions 57

of their larger counterparts. This lower sequence identity and size do not however preclude 58

the conservation of the elements of ordered secondary structure as well as the identity and 59

geometric arrangement of the residues configuring the active site (Figure 1). This structural 60

conservation can be even extended to the other non-toxic members of the larger fungal 61

extracellular RNases family, such as RNase T1 and RNase U2, for example [36, 37]. In 62

fact, comparison of all these RNases three-dimensional structures reveals the strict 63

conservation of the active site residues forming their catalytic triad, His50, Glu96, and 64

His137 in α-sarcin [38], for example, as well as the preservation of a highly hydrophobic 65

residue at the position corresponding to α-sarcin’s Leu145 [39] (Figure 1). On the other 66

hand, the presence of an Asp residue in HtA [40] and anisoplin [35] in a position equivalent 67

to α-sarcin Tyr48 [41] must be highlighted as a novelty in the active site of this family of 68

toxic RNases (Figure 1). This is a quite intriguing observation in the context that 69

substitution of this Tyr48 by Phe rendered a α-sarcin variant which was catalytically 70

incompetent, unable to inactivate the ribosome [41]. Interestingly, studies with HtA 71

mutants D40N and D40N/E66Q demonstrated an important role for Asp40 in the activity of 72

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

3

HtA in establishing a new set of electrostatic interactions different from the one described 73

for the already known larger ribotoxins [40]. 74

In the work herein presented α-sarcin Tyr 48 has been replaced by an Asp residue to 75

produce and characterize the corresponding Y48D mutant. As a control, the corresponding 76

inverse anisoplin mutant (D43Y) has been also studied. The results shown reveal the key 77

role of these residues not only in maintaining the correct electrostatic environment and 78

active site plasticity in each type of ribotoxins but also in the preservation of their 79

characteristic high thermostability. 80

MATERIALS AND METHODS 81

Mutant cDNA construction 82

All materials and reagents were of molecular biology grade. Cloning procedures, PCR-83

based oligonucleotide site-directed mutagenesis, and bacterial manipulations were carried 84

out as previously described [11, 16, 41-43]. Mutagenesis constructions were performed 85

using different sets of complementary mutagenic primers (Table S1). Mutations were 86

confirmed by DNA sequencing at the corresponding Complutense University facility. The 87

plasmids used as templates for mutagenesis, containing the cDNA sequence of either wild-88

type α-sarcin or anisoplin, have already been described [35, 38, 42]. 89

Protein production and purification 90

Escherichia coli RB791 or BL21 (DE3) cells, the latter ones being previously 91

cotransformed with a thioredoxin-producing plasmid (pT-Trx), and the corresponding wild-92

type or mutant plasmids were used to produce and purify all proteins from the periplasmic 93

soluble fraction, as previously described [35, 38, 42, 44-46]. The only exception was fungal 94

wild-type α-sarcin which was isolated from Aspergillus giganteus MDH18894, its natural 95

source, following the procedure reported before [11]. SDS-PAGE of proteins, Western 96

blots, protein hydrolysis, and amino acid analysis were performed according to standard 97

procedures, also as previously described [11, 42]. All four proteins studied were purified to 98

homogeneity according to their SDS-PAGE behavior and amino acid analysis. 99

Spectroscopic characterization 100

Spectroscopic characterization was performed following well established procedures [11, 101

22, 38, 41, 44, 47-51]. Absorbance measurements were carried out on a Shimadzu UV-102

1800 at 200 nm/min scanning speed and room temperature. Amino acid analyses and the 103

corresponding UV absorbance spectra were also used to calculate their extinction 104

coefficients (Table 1). Circular dichroism (CD) spectra were obtained in a Jasco 715 105

spectropolarimeter (Jasco, Easton, MD, USA), equipped with a thermostated cell holder 106

and a Neslab-111 circulating water bath, at 50 nm/min scanning speed. Thermal 107

denaturation profiles were recorded by measuring the temperature dependence of the 108

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

4

ellipticity at 220 nm in the 25 – 80ºC range using a rate of temperature increment of 30ºC 109

per hour. Fluorescence emission spectra were recorded on an SLM Aminco 8000 110

spectrofluorimeter at 25ºC using a slit width of 4 nm for both excitation and emission 111

beams. The spectra were recorded for excitation at 275 and 295 nm and both were 112

normalized by considering that Tyr emission above 380 nm is negligible. The Tyr 113

contribution was calculated as the difference between the two normalized spectra. 114

Thermostated cells with a path length of 0.2 and 1.0 cm for the excitation and emission 115

beams, respectively, were used. All these experiments were performed in 50 mM sodium 116

phosphate, pH 7.0. 117

Ribonucleolytic activity assays 118

The ribonucleolytic activity of ribotoxins on rabbit ribosomes was followed by detecting 119

the release of a specific 400-nt rRNA fragment, known as the α-fragment, from the 120

ribosomes of a cell-free rabbit reticulocyte lysate (Promega) as described [7, 21, 43, 52]. 121

Visualization of this α-fragment was performed by ethidium bromide staining of 2.0% 122

agarose gels after electrophoretic fractionation of the samples using denaturing conditions. 123

The protein concentration needed to produce 50% of cleavage of the 28S rRNA (IC50 124

value) was determined to evaluate proteins’ specific ribonucleolytic activity. 125

The specific cleavage by ribotoxins of a 35-mer synthetic oligonucleotide 126

mimicking the sequence and structure of the SRL was also analyzed as described before [7, 127

43, 53]. The sequence of this oligo was 5’-128

GGGAAUCCUGCUCAGUACGAGAGGAACCGCAGGUU -3’, where the cleavage site 129

by α-sarcin appears underlined. Synthesis of this SRL-like RNA oligo was performed as 130

previously described [7, 21]. Reaction products were run on a denaturing 19% (w/v) 131

polyacrylamide gel and visualized by ethidium bromide staining. 132

RESULTS 133

Spectroscopic characterization 134

Far-UV CD spectra of wild-type α-sarcin and the corresponding Y48D mutant showed 135

some differences but were still similar enough as to consider that both proteins displayed a 136

practically identical overall globular fold (Figure 2). Analysis using different software or 137

online tools has been proven to be not very useful in the particular case of ribotoxins given 138

their small size, their high degree of β-sheet content and, above all, the highly unusual 139

contribution of aromatic side-chains within this wavelength range [50, 54-56]. However, 140

the observed differences can be explained by minor local changes attributable to the 141

proximity of Trp51, an amino acid residue showing a non-negligible contribution in the far-142

UV wavelength region due to its involvement in a cation-π interaction with the ring of 143

His82 (Figure 3) [43, 50]. This interaction would be disturbed in the Y48D mutant. In 144

agreement with this hypothesis, Trp emission (Figure 4) was also shown to be more 145

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

5

heterogeneous, blue-shifted, and four-fold enhanced in the Y48D mutant (Table 1), as 146

expected from an increased hydrophobic microenvironment around the side-chain of Trp51, 147

upon disappearance of the mentioned cation-π interaction. In accordance with this 148

observation, and with the reported existence of tyrosine to Trp51 non-radiative energy 149

transfer in the wild-type protein [50], Tyr emission was also two-fold higher in the Y48D 150

mutant (Figure 4, Table 1), in spite of being a protein species with one Tyr residue less. 151

A similar situation took place when examining the far-UV CD features of wild-type 152

anisoplin and its D43Y variant; however differences were larger (Figure 2). The spectrum 153

of the wild-type protein corresponds to a protein with a high content of β-sheet, non-154

ordered structures and aromatic amino acid residues [26, 35, 55, 57], and the observed 155

changes can be attributed to the local contribution of the introduction of the new Tyr 156

residue within the active site of the protein (Figure 2). A global conformational change 157

cannot be however dismissed. Analysis of the fluorescence emission of wild-type anisoplin 158

upon excitation at 275 nm revealed the existence of a contribution centered at 320 nm 159

(Figure 5) which should not be attributable to Tyr residues because they barely emit around 160

this wavelength. This emission would rather be an indication of the already reported 161

different microenvironment surrounding anisoplin Trp residues [35]. This emission 162

disappears in the D43Y mutant, together with the appearance of a 4 nm red-shift of the Trp 163

spectra (from 333 to 337 nm, Figure 5). This set of results, including the far-UV CD 164

spectra, is therefore consistent with a relaxation of anisoplin conformational global fold 165

which exposes Trp side-chains to a less apolar environment. On the other hand Tyr 166

emission remains practically undetectable in spite of the introduction of a new Tyr residue 167

in this mutant. 168

Thermostability of both mutants 169

All four proteins studied showed thermograms compatible with the existence of a well-170

defined two-state thermal transition, confirming the adoption of a folded conformation 171

(Figure S1). Both wild-type α-sarcin and anisoplin are thermostable proteins with Tm values 172

of 52 [48, 54] and 61ºC [16], respectively (Table 1). Introduction of the Asp residue within 173

the α-sarcin active-site resulted in a dramatic reduction of the Tm value to 39ºC (Table 1). 174

Accordingly, the production yield of this mutant increased 15-fold when the E. coli cells 175

harboring the corresponding plasmid were grown at 25ºC instead of the standard 176

temperature of 37ºC (Table 1). This argument does not however justify the low yield 177

obtained for wild-type anisoplin (Table 1) which is not easy to explain given the present 178

knowledge in the ribotoxins’ field. It can be speculated that, in comparison to α-sarcin, 179

wild-type anisoplin would be more effective against prokaryotic than mammalian 180

ribosomes. This feature would be probably lost in the D43Y mutant after the introduction 181

of the Tyr residue within its active site (Table 1). Interestingly, this reverse mutation in 182

anisoplin introduced a dramatic stabilization of the protein, with a Tm value 14ºC higher 183

than the wild-type protein (Table 1). In agreement with this observation, the fluorescence 184

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

6

emission spectra of this mutant in the presence or the absence of 6 M urea were practically 185

indistinguishable (Figure 7) suggesting that the protein is still folded in the presence of high 186

concentrations of this denaturing agent. 187

Functional characterization 188

Ribotoxins are highly specific RNases against intact ribosomes, and they retain this 189

specificity when assayed against naked rRNA containing the SRL [1, 11, 21]. Therefore, 190

two different types of specific enzymatic assays are usually performed to measure their 191

enzymatic activity. The first, and most specific, is one that uses ribosomes within a rabbit 192

cell-free reticulocyte lysate [21]. Their highly specific cleavage can be then visualized by 193

detecting the release of a 400-nucleotide long rRNA fragment (the α-fragment) on a 194

denaturing agarose gel stained with ethidium bromide. The second assay frequently used is 195

based on the employment of short oligoribonucleotides mimicking the SRL sequence and 196

structure. Ribotoxins cleave specifically these SRL-like oligos, producing only two smaller 197

fragments which can be fractionated on a polyacrylamide gel [21, 58]. This cleavage is still 198

specific but several orders of magnitude less efficient than that produced on intact 199

ribosomes because it lacks important recognition determinants which are present at the 200

intact full ribosomes [20, 48, 49, 59-61]. 201

As thoroughly described before [1, 38, 42, 44, 51], wild-type α-sarcin was fully competent 202

against rabbit ribosomes showing an IC50 value of about 80 nM (Fig. 6). On the other hand, 203

α-sarcin Y48D was completely inactive by both criteria (Figure 6). This observation is in 204

agreement with previous studies where the removal of the hydroxyl group in the phenol 205

ring of Tyr48 rendered a catalytically inactive protein [41]. Tyr48 appears to be an essential 206

residue of α-sarcin’s active site. On the other hand, wild-type anisoplin showed less activity 207

than α-sarcin when assayed against intact ribosomes (IC50 = 930 nM), as described before 208

[35], whereas the D43Y mutant was at least as active (IC50 = 440 nM) as its wild-type 209

counterpart (Figure 7). However, this mutant also remained completely inactive when 210

assayed against the SRL-like oligomer (Figure 7), a substrate which lacks many of the 211

recognition regions needed for the specific catalytic action of ribotoxins. 212

DISCUSSION 213

The ribotoxins family has been thoroughly characterized over the past decades, focused 214

especially in α-sarcin and restrictocin, which are structurally very similar to non-toxic 215

ribonucleases (Figure 1). The discovery of smaller versions of these toxins, like HtA and 216

anisoplin, has prompted a change in the idea that all ribotoxins share a highly similar 217

structure. These "minimized" versions show identical specificity towards their substrate, the 218

SRL at the ribosome, but they share little amino-acid identity and have lost some structural 219

arrangements when compared to "canonical" ribotoxins like α-sarcin. In order to 220

understand these changes and how they may affect the functionality of ribotoxins, we have 221

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

7

produced different variants of α-sarcin and anisoplin where a key residue has been 222

interchanged. Particularly, Tyr48 in α-sarcin, which is essential for the catalytic activity of 223

α-sarcin [41] has been substituted for Asp, present in the smaller ribotoxins HtA and 224

anisoplin. Moreover, anisoplin Asp 3 has been also replaced for Tyr, to mimic α-sarcin 225

Tyr48. According to previous data, the native conformation of α-sarcin is preserved upon 226

conservative changes like the Y48F mutation [41]. In this new study the changes introduced 227

are more dramatic, involving charge changes that most likely will disturb the local 228

electrostatic arrangement, which is critical for the activity of the proteins studied [62-64]. 229

It has been reported how substitution of α-sarcin Trp51 by Phe results in 230

pronounced changes in the far-UV CD spectrum of the protein [50], highly resembling the 231

spectrum obtained now for the Y48D mutant (Figure 2A). This is explained because α-232

sarcin Trp51 displays a non-negligible contribution in the far-UV wavelength region due to 233

its involvement in a cation-π interaction with the ring of His82 (Figure 3) [43, 50]. As a 234

consequence of this interaction, Trp51 fluorescence emission is also practically 235

undetectable in the wild-type protein [50]. The far-UV CD spectrum of the Y48D α-sarcin 236

mutant suggests that this interaction is disturbed. In agreement with this hypothesis, Trp 237

emission was more heterogeneous (Figure 4) and four-fold enhanced in the Y48D mutant 238

(Table 1), suggesting that the disappearance of the cation-π interaction would increase the 239

hydrophobicity of the microenvironment around the side-chain of Trp51. This 240

interpretation agrees with a very similar observation, reported before, for another mutant 241

where His82 was the residue replaced (by Gln) making impossible the establishment of the 242

mentioned cation-π interaction [43]. Therefore, the far-UV CD small differences and Trp 243

emission changes observed can be explained by the local perturbation produced by the 244

removal of an aromatic moiety (Tyr48) in the spatial proximity of Trp51 (Figure 3) together 245

with the introduction of an additional negative charge. 246

The spectroscopic characterization of α-sarcin Y48D also revealed a 2-fold increase 247

in the Tyr emission (Figure 4, Table 1). In the wild-type protein, Tyr48 fluorescence 248

emission is completely quenched, and there is a Tyr to Trp51 non-radiative energy transfer 249

[41, 50]. By substituting Tyr for Asp the local configuration of Trp51 is altered, impeding 250

the energy transfer and therefore, increasing the fluorescence emission. 251

Detailed analysis of the spectroscopic features of the other mutant protein studied, 252

anisoplin D43Y, leads to different conclusions. First of all, the fluorescence emission of the 253

5 Tyr residues is barely detected, even in the wild-type protein (Figure 5). This emission 254

still goes undetected after the mutation, even though the change performed introduces a 255

new Tyr residue. On the other hand, the Trp population becomes more homogeneous and 256

solvent-exposed, at least in terms of their fluorescence emission behavior, as shown by the 257

red-shift in the spectrum (Figure 5). The far-UV CD changes observed (Figure 2) also point 258

towards this direction. The small size of the protein, altogether with its high content in β-259

sheet structure and non-ordered loops, results in a spectrum which is highly susceptible to 260

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

8

changes in the environment of its quite abundant Trp residues (Table 1). Overall, the results 261

obtained from the spectroscopic characterization of anisoplin D43Y suggest the induction 262

of local changes within the active site of the protein, as shown for the α-sarcin mutant, but 263

also a relaxation of the global protein fold which results in the mentioned homogenization 264

of the Trp residues microenvironment. 265

Replacement of Tyr48 by the negatively charged side-chain of an Asp residue 266

seems to strongly destabilize α-sarcin. In α-sarcin, Glu140 displays unusual backbone 267

torsional angles forming a salt bridge with Lys11 in the N-terminal β-hairpin (Figure 8) 268

[63]. This interaction, together with a well-defined network of hydrogen bonds, helps to 269

maintain the orientation of loop 5 [38, 51] and defines the required optimum hydrophobic 270

environment for enzymatic activity. Furthermore, His50, another key residue for activity 271

[38], also appears in the spatial vicinity of the mutated residue (Figure 3), whereas the 272

network of interactions involving Tyr106, Lys114 and Y48 is also essential for catalysis 273

(Figure 8) [43]. These interactions would be disrupted in the Y48D mutant. It appears that 274

the presence of an aromatic residue (Y48) within the active center would be critical for 275

maintaining the local interactions that render a high thermostable and specific RNase. This 276

idea is supported by the functional studies, which show that the α-sarcin Y48D mutant 277

completely loses its activity against rRNA specific substrates, such as ribosomes or a SRL-278

like oligomer (Figure 6). 279

A very different picture emerges, however, when a minimized ribotoxin such as 280

anisoplin is studied. Replacement of Asp43 by Tyr yielded an extremely heat resistant 281

variant (Table 1), which agrees with the loss of thermostability of α-sarcin Y48D mutant. 282

This extremely resistant protein did not display any ribonucleolytic activity when assayed 283

against the specific naked rRNA represented by the SRL-like substrate (Figure 7). This Asp 284

43 residue must be very important in an additional set of interactions that only appears in 285

minimized ribotoxins, just as suggested before for HtA while characterizing different 286

mutations of this Asp (D40N, D40N/E66Q) [40]. On the other hand, the now studied 287

anisoplin D43Y mutant was fully active when assayed against intact ribosomes, suggesting 288

that the conservation of not yet determined interactions other than those ones involved in 289

the specific recognition of the SRL sequence seem to be extremely important for substrate 290

recognition. This set of results further supports the proposal that the active site of these 291

minimized ribotoxins such as anisoplin or HtA would show a higher degree of plasticity 292

than their known larger counterparts [16], being able to accommodate electrostatic and 293

structural changes not suitable for the other previously characterized larger ribotoxins. 294

Ribotoxins are considered natural engineered proteins evolved from the non-toxic 295

ribonucleases [65]. The newly characterized "minimized" ribotoxins, like HtA or anisoplin, 296

represent a compromise among conformational freedom, stability, specificity and active site 297

plasticity that allows these proteins to accommodate the characteristic abilities of ribotoxins 298

into a shorter amino acid sequence. These smaller versions of ribotoxins may present new 299

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

9

and yet unexplored structural arrangements, being our study one of the first in this 300

direction. 301

Acknowledgements 302

This work was supported by project AE1/16-20695 from the Universidad Complutense 303

(Madrid, Spain). M.Ol. was recipient of a FPU predoctoral fellowship from the Spanish 304

Ministerio de Educación. L.G.-O. was a postdoctoral researcher of the PICATA program 305

from the Campus de Excelencia Internacional Moncloa. 306

Author contributions 307

MML, MOl, LGO, and DSG designed and performed the experiments and analyzed the 308

data. JL, MOñ, JGG and AMP designed and supervised the research and discussed the data. 309

All authors wrote the manuscript and approved its final version. 310

References 311

1. Lacadena, J., Álvarez-García, E., Carreras-Sangrà, N., Herrero-Galán, E., Alegre-312

Cebollada, J., García-Ortega, L., Oñaderra, M., Gavilanes, J. G. & Martínez-del-Pozo, A. 313

(2007) Fungal ribotoxins: molecular dissection of a family of natural killers, FEMS 314

Microbiol Rev. 31, 212-237. 315

2. Olombrada, M., Lázaro-Gorines, R., López-Rodríguez, J. C., Martínez-del-Pozo, A., 316

Oñaderra, M., Maestro-López, M., Lacadena, J., Gavilanes, J. G. & García-Ortega, L. 317

(2017) Fungal Ribotoxins: A Review of Potential Biotechnological Applications, Toxins. 9, 318

71. 319

3. Jennings, J. C., Olson, B. H., Roga, V., Junek, A. J. & Schuurmans, D. M. (1965) α-320

Sarcin, a New Antitumor Agent. II. Fermentation and Antitumor Spectrum, Appl 321

Microbiol. 13, 322-326. 322

4. Schindler, D. G. & Davies, J. E. (1977) Specific cleavage of ribosomal RNA caused by 323

α-sarcin, Nucleic Acids Res. 4, 1097-1110. 324

5. Nierhaus, K. H., Schilling-Bartetzko, S. & Twardowski, T. (1992) The two main states 325

of the elongating ribosome and the role of the α-sarcin stem-loop structure of 23S RNA, 326

Biochimie. 74, 403-10. 327

6. Brigotti, M., Rambelli, F., Zamboni, M., Montanaro, L. & Sperti, S. (1989) Effect of α-328

sarcin and ribosome-inactivating proteins on the interaction of elongation factors with 329

ribosomes, Biochem J. 257, 723-7. 330

7. García-Ortega, L., Álvarez-García, E., Gavilanes, J. G., Martínez-del-Pozo, A. & 331

Joseph, S. (2010) Cleavage of the sarcin-ricin loop of 23S rRNA differentially affects EF-G 332

and EF-Tu binding, Nucleic Acids Res. 38, 4108-19. 333

8. Olmo, N., Turnay, J., González de Buitrago, G., López de Silanes, I., Gavilanes, J. G. & 334

Lizarbe, M. A. (2001) Cytotoxic mechanism of the ribotoxin α-sarcin. Induction of cell 335

death via apoptosis, Eur J Biochem. 268, 2113-2123. 336

9. Oñaderra, M., Mancheño, J. M., Gasset, M., Lacadena, J., Schiavo, G., Martínez-del-337

Pozo, A. & Gavilanes, J. G. (1993) Translocation of α-sarcin across the lipid bilayer of 338

asolectin vesicles, Biochem J. 295, 221-225. 339

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

10

10. Gasset, M., Mancheño, J. M., Lacadena, J., Turnay, J., Olmo, N., Lizarbe, M. A., 340

Martínez-del-Pozo, A., Oñaderra, M. & Gavilanes, J. G. (1994) α-Sarcin, a ribosome-341

inactivating protein that translocates across the membrane of phospholipid vesicles, Curr 342

Top Pept Protein Res. 1, 99-104. 343

11. Martínez-Ruiz, A., García-Ortega, L., Kao, R., Lacadena, J., Oñaderra, M., Mancheño, 344

J. M., Davies, J., Martínez-del-Pozo, A. & Gavilanes, J. G. (2001) RNase U2 and α-sarcin: 345

A study of relationships, Methods Enzymol. 341, 335-351. 346

12. Fernández-Puentes, C. & Carrasco, L. (1980) Viral infection permeabilizes mammalian 347

cells to protein toxins, Cell. 20, 769-75. 348

13. Gasset, M., Oñaderra, M., Thomas, P. G. & Gavilanes, J. G. (1990) Fusion of 349

phospholipid vesicles produced by the anti-tumour protein α-sarcin, Biochem J. 265, 815-350

22. 351

14. Gasset, M., Martínez-del-Pozo, A., Oñaderra, M. & Gavilanes, J. G. (1989) Study of 352

the interaction between the antitumour protein α-sarcin and phospholipid vesicles, Biochem 353

J. 258, 569-75. 354

15. Masip, M., Lacadena, J., Mancheño, J. M., Oñaderra, M., Martínez-Ruiz, A., Martínez-355

del-Pozo, A. & Gavilanes, J. G. (2001) Arginine 121 is a crucial residue for the specific 356

cytotoxic activity of the ribotoxin α-sarcin, Eur J Biochem. 268, 6190-6196. 357

16. Olombrada, M., Garía-Ortega, L., Lacadena, J., Oñaderra, M., Gavilanes, J. G. & 358

Martínez-Del-Pozo, A. (2016) Involvement of loop 5 lysine residues and the N-terminal β-359

hairpin of the ribotoxin hirsutellin A on its insecticidal activity, Biol Chem. 397, 135-45. 360

17. Olombrada, M., Herrero-Galán, E., Tello, D., Oñaderra, M., Gavilanes, J. G., Martínez-361

del-Pozo, A. & García-Ortega, L. (2013) Fungal extracellular ribotoxins as insecticidal 362

agents, Insect Biochem Mol Biol. 43, 39-46. 363

18. Olombrada, M., Martínez-del-Pozo, A., Medina, P., Budia, F., Gavilanes, J. G. & 364

García-Ortega, L. (2014) Fungal ribotoxins: Natural protein-based weapons against insects, 365

Toxicon. 83, 69-74. 366

19. Arruda, L. K., Mann, B. J. & Chapman, M. D. (1992) Selective expression of a major 367

allergen and cytotoxin, Asp f I, in Aspergillus fumigatus. Implications for the 368

immunopathogenesis of Aspergillus-related diseases, J Immunol. 149, 3354-9. 369

20. Wool, I. G. (1997) Structure and mechanism of action of the cytotoxic ribonuclease α-370

sarcin in Ribonucleases (D'Alessio, G. & Riordan, J. F., eds) pp. 131-162, Academic Press 371

Inc, San Diego. 372

21. Kao, R., Martínez-Ruiz, A., Martínez-del-Pozo, A., Crameri, R. & Davies, J. (2001) 373

Mitogillin and related fungal ribotoxins, Methods Enzymol. 341, 324-335. 374

22. García-Ortega, L., Lacadena, J., Villalba, M., Rodríguez, R., Crespo, J. F., Rodríguez, 375

J., Pascual, C., Olmo, N., Oñaderra, M., Martínez-del-Pozo, A. & Gavilanes, J. G. (2005) 376

Production and characterization of a noncytotoxic deletion variant of the Aspergillus 377

fumigatus allergen Asp f 1 displaying reduced IgE binding, FEBS J. 272, 2536-44. 378

23. Boucias, D. G., Farmerie, W. G. & Pendland, J. C. (1998) Cloning and sequencing of 379

cDNA of the insecticidal toxin hirsutellin A, J Invertebr Pathol. 72, 258-61. 380

24. Herrero-Galán, E., Lacadena, J., Martínez-del-Pozo, A., Boucias, D. G., Olmo, N., 381

Oñaderra, M. & Gavilanes, J. G. (2008) The insecticidal protein hirsutellin A from the mite 382

fungal pathogen Hirsutella thompsonii is a ribotoxin, Proteins. 72, 217-228. 383

25. Maimala, S., Tartar, A., Boucias, D. & Chandrapatya, A. (2002) Detection of the toxin 384

Hirsutellin A from Hirsutella thompsonii, J Invertebr Pathol. 80, 112-26. 385

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

11

26. Viegas, A., Herrero-Galán, E., Oñaderra, M., Macedo, A. L. & Bruix, M. (2009) 386

Solution structure of hirsutellin A. New insights into the active site and interacting 387

interfaces of ribotoxins, FEBS J. 276, 2381-2390. 388

27. Lin, A., Huang, K. C., Hwu, L. & Tzean, S. S. (1995) Production of type II ribotoxins 389

by Aspergillus species and related fungi in Taiwan, Toxicon. 33, 105-110. 390

28. Parente, D., Raucci, G., Celano, B., Pacilli, A., Zanoni, L., Canevari, S., Adobati, E., 391

Colnaghi, M. L., Dosio, F., Arpicco, S., Cattel, L., Mele, A. & De Santis, R. (1996) Clavin 392

a type-1 ribosome-inactivating protein from Aspergillus clavatus IFO 8605 - cDNA 393

isolation, heterologous expression, biochemical and biological characterization of the 394

recombinant protein, Eur J Biochem. 239, 272-280. 395

29. Huang, K.-C., Hwang, Y.-Y., Hwang, L. & Lin, A. (1997) Characterization of a new 396

ribotoxin gene (c-sar) from Aspergillus clavatus, Toxicon. 35, 383-392. 397

30. Wirth, J., Martínez-del-Pozo, A., Mancheño, J. M., Martinez-Ruiz, A., Lacadena, J., 398

Oñaderra, M. & Gavilanes, J. G. (1997) Sequence determination and molecular 399

characterization of gigantin, a cytotoxic protein produced by the mould Aspergillus 400

giganteus IFO 5818, Arch Biochem Biophys. 343, 188-193. 401

31. Martínez-Ruiz, A., Kao, R., Davies, J. & Martínez-del-Pozo, A. (1999) Ribotoxins are 402

a more widespread group of proteins within the filamentous fungi than previously believed, 403

Toxicon. 37, 1549-1563. 404

32. Martínez-Ruiz, A., Martínez-del-Pozo, A., Lacadena, J., Oñaderra, M. & Gavilanes, J. 405

G. (1999) Hirsutellin A displays significant homology to microbial extracellular 406

ribonucleases, J Invertebr Pathol. 74, 96-97. 407

33. Herrero-Galán, E., García-Ortega, L., Olombrada, M., Lacadena, J., Martínez-del-408

Pozo, A., Gavilanes, J. G. & Oñaderra, M. (2013) Hirsutellin A: A Paradigmatic Example 409

of the Insecticidal Function of Fungal Ribotoxins, Insects. 4, 339-356. 410

34. Varga, J. & Samson, R. A. (2008) Ribotoxin genes in isolates of Aspergillus section 411

Clavati, Antonie Van Leeuwenhoek. 94, 481-485. 412

35. Olombrada, M., Medina, P., Budia, F., Gavilanes, J. G., Martínez-Del-Pozo, A. & 413

García-Ortega, L. (2017) Characterization of a new toxin from the entomopathogenic 414

fungus Metarhizium anisopliae: the ribotoxin anisoplin, Biol Chem. 398, 135-142. 415

36. Pace, C. N., Heinemann, U., Hahn, U. & Saenger, W. (1991) Ribonuclease T1: 416

Structure, function and stability, Angew Chem Int Ed Engl. 30, 343-360. 417

37. Noguchi, S., Satow, Y., Uchida, T., Sasaki, C. & Matsuzaki, T. (1995) Crystal 418

structure of Ustilago sphaerogena ribonuclease U2 at 1.8 angstrom resolution, 419

Biochemistry. 34, 15583-15591. 420

38. Lacadena, J., Martínez-del-Pozo, A., Martínez-Ruiz, A., Pérez-Cañadillas, J. M., Bruix, 421

M., Mancheño, J. M., Oñaderra, M. & Gavilanes, J. G. (1999) Role of histidine-50, 422

glutamic acid-96, and histidine-137 in the ribonucleolytic mechanism of the ribotoxin α-423

sarcin, Proteins. 37, 474-484. 424

39. Masip, M., García-Ortega, L., Olmo, N., García-Mayoral, M. F., Pérez-Cañadillas, J. 425

M., Bruix, M., Oñaderra, M., Martínez-del-Pozo, A. & Gavilanes, J. G. (2003) Leucine 145 426

of the ribotoxin α-sarcin plays a key role for determining the specificity of the ribosome-427

inactivating activity of the protein, Prot Science. 12, 161-169. 428

40. Herrero-Galán, E., García-Ortega, L., Lacadena, J., Martínez-Del-Pozo, A., Olmo, N., 429

Gavilanes, J. G. & Oñaderra, M. (2012) Implication of an Asp residue in the 430

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

12

ribonucleolytic activity of hirsutellin A reveals new electrostatic interactions at the active 431

site of ribotoxins, Biochimie. 94, 427-433. 432

41. Álvarez-García, E., García-Ortega, L., Verdun, Y., Bruix, M., Martínez-del-Pozo, A. & 433

Gavilanes, J. G. (2006) Tyr-48, a conserved residue in ribotoxins, is involved in the RNA-434

degrading activity of α-sarcin, Biol Chem. 387, 535-41. 435

42. Lacadena, J., Martínez-del-Pozo, A., Barbero, J. L., Mancheño, J. M., Gasset, M., 436

Oñaderra, M., López-Otín, C., Ortega, S., García, J. L. & Gavilanes, J. G. (1994) 437

Overproduction and purification of biologically active native fungal α-sarcin in 438

Escherichia coli, Gene. 142, 147-151. 439

43. Castaño-Rodríguez, C., Olombrada, M., Partida-Hanon, A., Lacadena, J., Oñaderra, 440

M., Gavilanes, J. G., García-Ortega, L. & Martínez-del-Pozo, A. (2015) Involvement of 441

loops 2 and 3 of α-sarcin on its ribotoxic activity, Toxicon. 96, 1-9. 442

44. García-Ortega, L., Lacadena, J., Lacadena, V., Masip, M., de Antonio, C., Martínez-443

Ruiz, A. & Martínez-del-Pozo, A. (2000) The solubility of the ribotoxin α-sarcin, produced 444

as a recombinant protein in Escherichia coli, is increased in the presence of thioredoxin, 445

Lett Appl Microbiol. 30, 298-302. 446

45. Mancheño, J. M., Gasset, M., Albar, J. P., Lacadena, J., Martinez-del-Pozo, A., 447

Oñaderra, M. & Gavilanes, J. G. (1995) Membrane interaction of a β-structure-forming 448

synthetic peptide comprising the 116-139th sequence region of the cytotoxic protein α-449

sarcin, Biophys J. 68, 2387-95. 450

46. Siemer, A., Masip, M., Carreras, N., García-Ortega, L., Oñaderra, M., Bruix, M., 451

Martínez del Pozo, A. & Gavilanes, J. G. (2004) Conserved asparagine residue 54 of α-452

sarcin plays a role in protein stability and enzyme activity, Biol Chem. 385, 1165-70. 453

47. Álvarez-García, E., Martínez-del-Pozo, A. & Gavilanes, J. G. (2009) Role of the basic 454

character of α-sarcin's NH2-terminal β-hairpin in ribosome recognition and phospholipid 455

interaction, Arch Biochem Biophys. 481, 37-44. 456

48. García-Ortega, L., Lacadena, J., Mancheño, J. M., Oñaderra, M., Kao, R., Davies, J., 457

Olmo, N., Martínez-del-Pozo, A. & Gavilanes, J. G. (2001) Involvement of the amino-458

terminal β-hairpin of the Aspergillus ribotoxins on the interaction with membranes and 459

nonspecific ribonuclease activity, Protein Sci. 10, 1658-68. 460

49. García-Ortega, L., Masip, M., Mancheño, J. M., Oñaderra, M., Lizarbe, M. A., García-461

Mayoral, M. F., Bruix, M., Martínez-del-Pozo, A. & Gavilanes, J. G. (2002) Deletion of the 462

NH2-terminal β-hairpin of the ribotoxin α-sarcin produces a nontoxic but active 463

ribonuclease, J Biol Chem. 277, 18632-9. 464

50. De Antonio, C., Martínez-del-Pozo, A., Mancheño, J. M., Oñaderra, M., Lacadena, J., 465

Martínez-Ruiz, A., Pérez-Cañadillas, J. M., Bruix, M. & Gavilanes, J. G. (2000) 466

Assignment of the contribution of the tryptophan residues to the spectroscopic and 467

functional properties of the ribotoxin α-sarcin, Proteins. 41, 350-61. 468

51. Lacadena, J., Mancheño, J. M., Martínez-Ruiz, A., Martínez-del-Pozo, A., Gasset, M., 469

Oñaderra, M. & Gavilanes, J. G. (1995) Substitution of histidine-137 by glutamine 470

abolishes the catalytic activity of the ribosome-inactivating protein α-sarcin, Biochem J. 471

309, 581-586. 472

52. Herrero-Galán, E., García-Ortega, L., Lacadena, J., Martínez-Del-Pozo, A., Olmo, N., 473

Gavilanes, J. G. & Oñaderra, M. (2012) A non-cytotoxic but ribonucleolytically specific 474

ribotoxin variant: implication of tryptophan residues in the cytotoxicity of hirsutellin A, 475

Biol Chem. 393, 449-456. 476

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

13

53. Olombrada, M., Rodríguez-Mateos, M., Prieto, D., Pla, J., Remacha, M., Martínez-del-477

Pozo, A., Gavilanes, J. G., Ballesta, J. P. & García-Ortega, L. (2014) The acidic ribosomal 478

stalk proteins are not required for the highly specific inactivation exerted by α-sarcin of the 479

eukaryotic ribosome, Biochemistry. 53, 1545-1547. 480

54. Martínez-del-Pozo, A., Gasset, M., Oñaderra, M. & Gavilanes, J. G. (1988) 481

Conformational study of the antitumor protein α-sarcin, Biochim Biophys Acta. 953, 280-482

288. 483

55. Lacadena, J., Martínez-del-Pozo, A., Gasset, M., Patiño, B., Campos-Olivas, R., 484

Vázquez, C., Martínez-Ruiz, A., Mancheño, J. M., Oñaderra, M. & Gavilanes, J. G. (1995) 485

Characterization of the antifungal protein secreted by the mould Aspergillus giganteus, 486

Arch Biochem Biophys. 324, 273-81. 487

56. Woody, R. W. a. D., A.K. (1996) Aromatica and cysteine side-chain CD in proteins, 488

Circular dichroism and Conformational Analysis of Biomolecules, 109-175. 489

57. Woody, R. W. (1978) Aromatic side-chain contributions to the far ultraviolet circular 490

dichroism of peptides and proteins, Biopolymers. 17, 1451-1467. 491

58. Endo, Y., Chan, Y.-L., Lin, A., Tsurugi, K. & Wool, I. (1988) The cytotoxins α-sarcin 492

and ricin retain their specificity when tested on a synthetic oligoribonucleotide (35-mer) 493

that mimics a region of 28 S ribosomal ribonucleic acid., J Biol Chem. 263, 7917-7920. 494

59. Glück, A. & Wool, I. G. (1996) Determination of the 28 S ribosomal RNA identity 495

element (G4319) for α-sarcin and the relationship of recognition to the selection of the 496

catalytic site, Journal of Molecular Biology. 256, 838-848. 497

60. García-Mayoral, F., García-Ortega, L., Álvarez-García, E., Bruix, M., Gavilanes, J. G. 498

& Martínez-del-Pozo, A. (2005) Modeling the highly specific ribotoxin recognition of 499

ribosomes, FEBS Lett. 579, 6859-64. 500

61. García-Mayoral, M. F., García-Ortega, L., Lillo, M. P., Santoro, J., Martínez-del-Pozo, 501

A., Gavilanes, J. G., Rico, M. & Bruix, M. (2004) NMR structure of the noncytotoxic α-502

sarcin mutant ∆(7-22): the importance of the native conformation of peripheral loops for 503

activity, Protein Sci. 13, 1000-1011. 504

62. Pérez-Cañadillas, J. M., Campos-Olivas, R., Lacadena, J., Martínez-del-Pozo, A., 505

Gavilanes, J. G., Santoro, J., Rico, M. & Bruix, M. (1998) Characterization of pKa values 506

and titration shifts in the cytotoxic ribonuclease α-sarcin by NMR. Relationship between 507

electrostatic interactions, structure, and catalytic function, Biochemistry. 37, 15865-15876. 508

63. Pérez-Cañadillas, J. M., Santoro, J., Campos-Olivas, R., Lacadena, J., Martínez-del-509

Pozo, A., Gavilanes, J. G., Rico, M. & Bruix, M. (2000) The highly refined solution 510

structure of the cytotoxic ribonuclease α-sarcin reveals the structural requirements for 511

substrate recognition and ribonucleolytic activity, J Mol Biol. 299, 1061-1073. 512

64. Campos-Olivas, R., Bruix, M., Santoro, J., Martínez-del-Pozo, A., Lacadena, J., 513

Gavilanes, J. G. & Rico, M. (1996) Structural basis for the catalytic mechanism and 514

substrate specificity of the ribonuclease α-sarcin, FEBS Lett. 399, 163-165. 515

65. Kao, R. & Davies, J. (1999) Molecular dissection of mitogillin reveals that the fungal 516

ribotoxins are a family of natural genetically engineered ribonucleases, J Biol Chem. 274, 517

12576-12582. 518

66. Martinez-Oyanedel, J., Choe, H. W., Heinemann, U. & Saenger, W. (1991) 519

Ribonuclease T1 with free recognition and catalytic site: crystal structure analysis at 1.5 A 520

resolution, J Mol Biol. 222, 335-52. 521

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

14

67. Noguchi, S. (2010) Structural changes induced by the deamidation and isomerization 522

of asparagine revealed by the crystal structure of Ustilago sphaerogena ribonuclease U2B, 523

Biopolymers. 93, 1003-10. 524

68. Noguchi, S. (2010) Isomerization mechanism of aspartate to isoaspartate implied by 525

structures of Ustilago sphaerogena ribonuclease U2 complexed with adenosine 3'-526

monophosphate, Acta Crystallogr D Biol Crystallogr. 66, 843-9. 527

69. DeLano, W. L. (2008) The PyMOL Molecular Graphics System, San Diego, 528

California. 529

530

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

15

Table 1. Purification yield, expressed as milligrams of protein isolated per one liter of 531

original culture grown at 37ºC, and spectroscopic parameters of the wild-type and mutant 532

proteins studied. Relative fluorescence emission yields (QTrp and QTyr) are referred to the 533

values of the wild-type proteins. Tm values are also shown. 534

Protein Purification

yielda) E0.1% b) Number

of Tyr Number of Trp QTrp QTyr Tm (ºC)

α-Sarcin WT

7.0 1.34 8 2 1.00 1.00 52

α-Sarcin Y48D

0.1c) 1.24 7 2 4.08 2.35 39

Anisoplin WT

0.5 1.62 5 3 1.00 - 61

Anisoplin D43Y

6.2 1.44 6 3 0.92 - 75

a) Expressed as mg per liter of original bacterial culture. 535

b)E0.1% (280 nm, 1.0 cm). 536

c)This yield increased to 1.5 mg when cells were grown at 25ºC in accordance with the 537

highly diminished Tm value of this mutant. 538

539

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

16

540

Figure 1: Representation of the three-dimensional structure and active center 541

geometric arrangement of representative fungal RNases. (A) Diagrams showing the 542

three-dimensional structure of ribotoxins α-sarcin (PDB ID: 1DE3) [63] and HtA (PDB ID: 543

2KAA) [26], and two non-toxic fungal extracellular RNases from the same family: RNases 544

T1 (9RNT) [36, 66] and U2 (1RTU) [37, 67, 68]. (B) Geometric arrangement of the active 545

site residues of these same four RNases. The catalytic triad made of two His and one Glu 546

residues is conserved in all proteins shown while a fourth residue, the equivalent to α-sarcin 547

Leu145, maintains its highly hydrophobic character (Phe or Leu). The position 548

corresponding to α-sarcin Tyr48 is also conserved except for HtA and anisoplin (not 549

shown) where the equivalent position is occupied by an Asp residue (D40, in red). 550

Diagrams were generated using the PyMol software [69]. 551

552

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

17

553

Figure 2. Far-UV CD spectra of the ribotoxins studied. Wild-type α-sarcin (A) and 554

anisoplin (B) ribotoxins (black lines) and their corresponding Y48D (A) and D43Y (B) 555

mutant variants (blue lines). Results are shown as mean residue weight ellipticity ([θ]MRW) 556

values expressed in units of degree x cm2 x dmol-1. 557

558

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

18

559

Figure 3. Diagram showing the relative spatial positions of Tyr48, His50, Trp51, and 560

His82 residues in α-sarcin. There is a cation-π interaction between Trp51 and His82 [50, 561

62, 63]. Diagram was generated using the PyMol software [69] and the corresponding PDB 562

coordinates (PDB ID: 1DE3). 563

564

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

19

565

Figure 4. Fluorescence emission spectra of wild-type α-sarcin and its Y48D mutant. 566

All spectra were recorded at identical protein concentrations. Spectra were recorded at 567

excitation wavelengths of 275 nm (continuous black line) and 295 nm (continuous blue 568

line). These two spectra were normalized at wavelengths above 380 nm to obtain the 569

Tryptophan contribution (dashed red line). Tyrosine contribution (dashed green line) was 570

calculated by subtracting the Trp only contribution (dashed red line) from that spectrum 571

obtained after excitation at 275 nm (continuous black line). Fluorescence emission units 572

were arbitrary, and referred to the maximum value of wild-type α-sarcin upon excitation at 573

275 nm. The positions of the two maxima corresponding to both spectra of the wild-type 574

and Y48D proteins upon excitation at 295 nm are indicated with a vertical black line. 575

576

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

20

577

578

Figure 5. Fluorescence emission spectra of wild-type anisoplin and its D43Y mutant. 579

(Upper panel) All spectra were recorded at identical protein concentrations. Spectra were 580

recorded at excitation wavelengths of 275 nm (continuous black line) and 295 nm 581

(continuous blue line). These two spectra were normalized at wavelengths above 380 nm to 582

obtain the Tryptophan contribution (dashed red line). Tyrosine contribution (dashed green 583

line) was calculated by subtracting the Trp only contribution (dashed red line) from that 584

spectrum obtained after excitation at 275 nm (continuous black line). Fluorescence 585

emission units were arbitrary, and referred to the maximum value of wild-type α-sarcin 586

upon excitation at 275 nm. The positions of the two maxima corresponding to both spectra 587

of the wild-type and D43Y proteins upon excitation at 295 nm are indicated with a vertical 588

black line. (Lower panel) Fluorescence emission spectra of the anisoplin D43Y mutant, 589

upon excitation at 275 nm, in the absence (black line) or in the presence of 6 M urea (red 590

line). 591

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

21

592

Figure 6. Specific ribonucleolytic activity of wild-type α-sarcin and its Y48D mutant. 593

(Upper panel) Ribosome cleaving activity assay performed using a rabbit cell-free 594

reticulocytes lysate. A control in the absence of enzyme is also shown (C-). Protein 595

concentrations shown are 20, 50, and 100 nM. The highly specific ribonucleolytic activity 596

of the ribotoxins is shown by the release of the 400-nt α-fragment (α) from the 28S rRNA 597

of eukaryotic ribosomes. Positions of bands corresponding to 28S, 18S, and 5S rRNA are 598

also indicated. The graph shows the quantitation of two independent experiments with error 599

bars representing ±SEM values. (Lower panel) Activity assay on a 35-mer oligonucleotide 600

mimicking the SRL. A control in the absence of enzyme is also shown (C-). Protein 601

concentrations shown are 100, 200, and 500 nM. The 21-mer and 14-mer oligonucleotides 602

resulting from the specific cleavage of a single phosphodiester bond, as well as the intact 603

35-mer oligo are indicated. 604

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

22

605

Figure 7. Specific ribonucleolytic activity of wild-type anisoplin and its D43Y mutant. 606

(Upper panel) Ribosome cleaving activity assay performed using a rabbit cell-free 607

reticulocytes lysate. A control in the absence of enzyme is also shown (C-). Protein 608

concentrations shown are 2.5, 25, and 250 nM. The highly specific ribonucleolytic activity 609

of the ribotoxins is shown by the release of the 400-nt α-fragment (α) from the 28S rRNA 610

of eukaryotic ribosomes. Positions of bands corresponding to 28S, 18S, and 5S rRNA are 611

also indicated. The graph shows the quantitation of two independent experiments with error 612

bars representing ±SEM values. (Lower panel) Activity assay on a 35-mer oligonucleotide 613

mimicking the SRL. A control in the absence of enzyme is also shown (C-). Protein 614

concentrations shown are 10, 50, and 250 nM. The 21-mer and 14-mer oligonucleotides 615

resulting from the specific cleavage of a single phosphodiester bond, as well as the intact 616

35-mer oligo are indicated. 617

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

23

618

Figure 8. Diagrams showing the relative spatial positions of different amino acid 619

residues in α-sarcin: (A) Lys11, Glu140, and His137; (B) Tyr48, Tyr106 and Lys114. 620

Diagrams were generated using the PyMol software [69] and the corresponding PDB 621

coordinates (PDB ID: 1DE3). 622

623